ANN Structures Research Articles

Developing an intelligent prediction system for successful aging based on artificial neural networks.

Due to the growing number of disabilities in elderly, Attention to this period of life is essential to be considered. Few studies focused on the physical, mental, disabilities, and disorders affecting the quality of life in elderly people. SA1 is related to various factors influencing the elderly's life. So, the objective of the current study is to build an intelligent system for SA prediction through ANN2 algorithms to investigate better all factors affecting the elderly life and promote them. This study was performed on 1156 SA and non-SA cases. We applied statistical feature reduction method to obtain the best factors predicting the SA. Two models of ANNs with 5, 10, 15, and 20 neurons in hidden layers were used for model construction. Finally, the best ANN configuration was obtained for predicting the SA using sensitivity, specificity, accuracy, and cross-entropy loss function. The study showed that 25 factors correlated with SA at the statistical level of P < 0.05. Assessing all ANN structures resulted in FF-BP3 algorithm having the configuration of 25-15-1 with accuracy-train of 0.92, accuracy-test of 0.86, and accuracy-validation of 0.87 gaining the best performance over other ANN algorithms. Developing the CDSS for predicting SA has crucial role to effectively inform geriatrics and health care policymakers decision making.

Investigation of Hydrodynamics Characteristics of Fluidized Bed with back propagation Artificial Neural Network (BPANN)

Correlations have also been developed with system parameters by using dimensional analysis and an artificial neural network approach. The paper describes an investigation for the thermal design of a fluidized bed cooler and prediction of heat transfer rate among the media categories. It is devoted to the thermal design of such equipment and their application in the industrial fields. In the present work, an extensive ANN by using back propagation (BP) has been carried out to correlate the expansion ratio, fluctuation ratio in gas-solid fluidized bed. Back propagation network is the most well-known and widely used among the current types of neural network system, several applications of ANN for modeling of nonlinear process systems and subsequent control were reported. In back-propagation, different ANN structures (I×H×O) with varying number of neurons in the hidden layer were used as a tool for training input and output data for prediction value of hydrodynamic characteristics of the bed system. It was noted that ring models are the best ones in reducing bed expansion ratio and fluctuation ratio around (25%) and (23.22%).

Assessing Feed-Forward Backpropagation Artificial Neural Networks for Strain-Rate-Sensitive Mechanical Modeling.

The manufacturing processes and design of metal and alloy products can be performed over a wide range of strain rates and temperatures. To design and optimize these processes using computational mechanics tools, the selection and calibration of the constitutive models is critical. In the case of hazardous and explosive impact loads, it is not always possible to test material properties. For this purpose, this paper assesses the efficiency and the accuracy of different architectures of ANNs for the identification of the Johnson-Cook material model parameters. The implemented computational tool of an ANN-based parameter identification strategy provides adequate results in a range of strain rates required for general manufacturing and product design applications. Four ANN architectures are studied to find the most suitable configuration for a reduced amount of experimental data, particularly for cases where high-impact testing is constrained. The different ANN structures are evaluated based on the model's predictive capability, revealing that the perceptron-based network of 66 inputs and one hidden layer of 30 neurons provides the highest prediction accuracy of the effective flow stress-strain behavior of Ti64 alloy and three virtual materials.

Analysis and Modelling of Temperature at the Water – Atmosphere Interface of a Lake by Energy Budget and ANNs Models

Different processes of lake surface water temperature (Tsw) are considered in this work. The Tsw estimation is based on the energy budget method and Artificial Neural Networks (ANNs models).These processes were applied to Vegoritis Lake in northern Greece. For the analysis, daily meteorological data, lake characteristics, as well as simulation results from an integrated heat transfer model were used for two complete years. The simulation results from the heat transfer model are considered as the reference and more accurate procedure to estimate Tsw. These results are used to compare the performance of the other processes. The examined processes include:(a) models of heat storage changes in relation to net radiation values Qt(Rn); (b) net radiation estimation using different approaches, such as the process of Slob’s equation with adjusted coefficients to lake data; and (c) ANN models with various architectures and input variables. The results show that the surface water energy budget model accurately describes the temperature (r2 = 0.916, RMSE = 2.422 °C). The ANN(5,6,1) model in which Tsw(i-1) is incorporated in the input variables was considered the best performing compared to all other ANN structures (r2 = 0.995, RMSE = 0.490 °C). Using different approaches for simulating net radiation (Rn) and Qt(Rn) in the equation of surface water temperature estimation provides results with lower accuracy.

Prediction of the Numbers of Visitors at the Sinop Museums by Artificial Neural Networks

In this study, the numbers of museums ‘visitors (Archaeology, Ethnography and Historical Prison) at the city center of Sinop province have been predicted by Artificial Neural Network structures. Artificial Neural Network models have been created in MATLAB environment. These Artificial Neural Network models are feed forward and trained by Backpropagation Algorithm. For each museum, a Artificial Neural Network with 19-inputs and 1-output have been used separately. As inputs of networks, 10 different meteorological factors, time factor (month, year), tourism income (TL), exchange rate ($/TL) and monthly-yearly PPI and CPI data have been used. Output of ANNs is the daily average of number of visitors for each month. In order to train and test the Artificial Neural Networks, the number of visitors of museum at city center for total 60 months of years between 2012 and 2017, and other input data have been used. The selection of proper Artificial Neural Networks structure have been achieved by trying backpropagation training functions 50-times on 3-different activation functions structure with 8 different neuron counts at one hidden layer. Totally, 32400-network have been created by training and the best network structure for each museum have been selected. Estimation result obtained by the Artificial Neural Network models have been evaluated and discussed. As a result of this work, it has been proved that estimation of number of visitors visiting museums at Sinop province can be done by using ANN structures.

DANNP: an efficient artificial neural network pruning tool

Background Artificial neural networks (ANNs) are a robust class of machine learning models and are a frequent choice for solving classification problems. However, determining the structure of the ANNs is not trivial as a large number of weights (connection links) may lead to overfitting the training data. Although several ANN pruning algorithms have been proposed for the simplification of ANNs, these algorithms are not able to efficiently cope with intricate ANN structures required for complex classification problems. Methods We developed DANNP, a web-based tool, that implements parallelized versions of several ANN pruning algorithms. The DANNP tool uses a modified version of the Fast Compressed Neural Network software implemented in C++ to considerably enhance the running time of the ANN pruning algorithms we implemented. In addition to the performance evaluation of the pruned ANNs, we systematically compared the set of features that remained in the pruned ANN with those obtained by different state-of-the-art feature selection (FS) methods. Results Although the ANN pruning algorithms are not entirely parallelizable, DANNP was able to speed up the ANN pruning up to eight times on a 32-core machine, compared to the serial implementations. To assess the impact of the ANN pruning by DANNP tool, we used 16 datasets from different domains. In eight out of the 16 datasets, DANNP significantly reduced the number of weights by 70%–99%, while maintaining a competitive or better model performance compared to the unpruned ANN. Finally, we used a naïve Bayes classifier derived with the features selected as a byproduct of the ANN pruning and demonstrated that its accuracy is comparable to those obtained by the classifiers trained with the features selected by several state-of-the-art FS methods. The FS ranking methodology proposed in this study allows the users to identify the most discriminant features of the problem at hand. To the best of our knowledge, DANNP (publicly available at www.cbrc.kaust.edu.sa/dannp) is the only available and on-line accessible tool that provides multiple parallelized ANN pruning options. Datasets and DANNP code can be obtained at www.cbrc.kaust.edu.sa/dannp/data.php and https://doi.org/10.5281/zenodo.1001086.

Urban sprawl assessment and modeling using landsat images and GIS

Nowadays, urban sprawl phenomenon have been seen in many of cities in the developing and developed countries. Urban sprawl is considered as a particular kind of urban growth which comes up with a lot of negative effects. Thus, monitoring, analyzing and modeling of this phenomenon seem to be unavoidable. This paper assess urban sprawl in Tehran Metropolis as the capital of Iran and models urban sprawl in this mega city utilizing artificial neural networks and adaptive neuro-based fuzzy inference system methods with remote sensing data and geospatial information systems spatial analyses and modeling capabilities to simulate Tehran urban growth. The results confirm that this city has experienced sprawl and sprawl has an increasing rate. Three Landsat imageries from TM and ETM+ sensors taken in 1988, 1999 and 2010 and seven predictor variables include distance to road, distance to green space, slope, elevation, distance to fault, distance to developed area and number of urban cells in 3 by 3 neighborhoods have been used for urban sprawl assessment and modeling. Relative operating characteristics (ROC) and sensitivity analyses have been used to evaluate simulation results. In this research, two evaluation steps have been implemented using ROC and on the both of them, ANFIS presented the best performance vs two different proposed ANN structures.

Efficient neural network modeling of reconfigurable microstrip patch antenna through knowledge‐based three‐step strategy

SummaryArtificial neural network (ANN) provides an efficient modeling technique based on input–output data obtained from an engineering problem. Highly nonlinear and complex relationships can be formed by ANN because of its nonlinear nature and representing knowledge at interconnection weights. If an ANN model is not sufficient in respect to the accuracy and time consumption, the knowledge based on design experience can be embedded into the modeling process. This knowledge reduces the complexity of the nonlinear input–output relationships; therefore, the knowledge embedded ANN can be formed easily compared with the conventional ANN model. Three‐step modeling strategy generates the initial knowledge via the conventional ANN modeling and consists of three sequential and also discrete training processes exploiting the knowledge‐based methods such as prior knowledge input and prior knowledge input with difference. The latter step improves the accuracy of the former step, and three‐step modeling provides more accuracy than conventional ANN modeling. The efficiency of three‐step modeling strategy is demonstrated on two data sets, which are obtained by the reconfigurable microstrip patch antenna design problem. The different number of neurons and ANN structures is handled for the comparison as well. In addition, ANN modeling is formed by MATLAB m‐file through neural network toolbox to reveal the efficiency of knowledge‐based three‐step modeling strategy. Copyright © 2016 John Wiley &amp; Sons, Ltd.

Application of Artificial Neural Network and Adaptive Neural-based Fuzzy Inference System Techniques in Estimating of Virtual Water

Wheat, barley, sugerbeet, potato, alfalfa, and corn are common crops produced in Iran, which need the most virtual water volume compared to other crops. Determination of the virtual water for these crops would assist in better management of water resources. The main objective of this study is to find out the best technique for estimating and mapping of virtual water. In this research, the virtual water volume was determined by crop water requirement and crop yields using three ANN structures as well as ANFIS technique. Based on RMSE and R the comparison of obtained results predicted through the applied ANNs structures indicate that the RBF outperforms the other models for estimating virtual water for wheat, potato, corn, and barley. Moreover, a comparison between RBF and ANFIS revealed that ANFIS is a promising model, which can be efficient mathematical tool for estimation of crop‟s virtual water.

Fuzzy wavelet network identification of optimum operating point of non-crystalline silicon solar cells

The emerging non-crystalline silicon (c-Si) solar cell technologies are starting to make significant inroads into solar cell markets. Most of the researchers have focused on c-Si solar cell in maximum power points tracking applications of photovoltaic (PV) systems. However, the characteristics of non-c-Si solar cell technologies at maximum power point (MPP) have different trends in current–voltage characteristics. For this reason, determining the optimum operating point is very important for different solar cell technologies to increase the efficiency of PV systems. In this paper, it has been shown that the use of fuzzy system coupled with a discrete wavelet network in Takagi–Sugeno type model structure is capable of identifying the MPP voltage of different non-c-Si solar cells with very high accuracy. The performance of the fuzzy-wavelet network (FWN) method has been compared with other ANN structures, such as radial basis function (RBF), adaptive neuro-fuzzy inference system (ANFIS) and three layered feed-forward neural network (TFFN). The simulation results show that the single FWN architecture has superior approximation accuracy over the other methods and a very good generalization capability for different operating conditions and different technologies.

The role of stepover ratio in prediction of surface roughness in flat end milling

Surface roughness prediction studies in end milling operations are usually based on three main parameters composed of cutting speed, feed rate and depth of cut. The stepover ratio is usually neglected without investigating it. The aim of this study is to discover the role of the stepover ratio in surface roughness prediction studies in flat end milling operations. In realising this, machining experiments are performed under various cutting conditions by using sample specimens. The surface roughnesses of these specimens are measured. Two ANN structures were constructed. First of them was arranged with considering, and the second without considering the stepover ratio. ANN structures were trained and tested by using the measured data for predicting the surface roughness. Average RMS error of the ANN model considering stepover ratio is 0.04 and without considering stepover ratio is 0.26. The first model proved capable of prediction of average surface roughness (Ra) with a good accuracy and the second model revealed remarkable deviations from the experimental values.

A Novel Approach for Protection and Condition Monitoring of Power Transformer Using MRBFNN

This article presents differential relaying and condition monitoring of power transformer using minimal radial basis function neural network (MRBFNN). This type of RBF neural network uses a sequential learning procedure to determine the optimum number of neurons in the hidden layer without resorting to trial and error. This ANN-based scheme monitors operating conditions of the transformer and detects the fault and issues the trip signal in case of an internal fault. It has been realized through two different ANN structures using the minimal radial basis function (MRBF) learning algorithm. The proposed protection scheme has been evaluated using simulated data obtained through EMTP/ATP package. The results amply demonstrate the capabilities of the proposed fault detector (FD) and condition monitor (CM) in terms of accuracy and speed with respect to detection of fault, classification and pattern recognition of different events of power transformer. The number of training patterns and training time are drastically reduced and significant accuracy is achieved.

Risk assessment due to local demand forecast uncertainty in the competitive supply industry

A risk assessment on local demand forecast uncertainty is presented. The aim is to highlight high-risk periods over different lengths of time and daily value-at-risk (VAR) due to load forecast errors. A number of load forecasts have been performed, and the load forecast is based on ARIMA models and ANN structures. With the residuals from load forecasting, the risk indexes over different time periods and seasons are formed. Moreover, a new methodology using the standard deviation of load increment on evaluating the risk is proposed. In contrast with the standard forecasting method that relies on a sophisticated forecast procedure, the new approach provides a useful and fast method to evaluate the risk due to load forecast uncertainty for a variety of local demand profiles. Finally, the VAR methodology combined with the NETA system is applied to a local electricity supplier in the UK.

Protection and Conditions Monitoring of Power Transformer Using ANN

This paper presents differential protection and condition monitoring scheme for power transformer using ANN. This ANN-based scheme monitors different operating conditions of the transformer, detects the fault, and issues the trip signal in case of internal fault only. The proposed protection scheme is accurate, fast, efficient, and reliable. This scheme has been realized through two different ANN structures using the Radial Basis Function (RBF) learning algorithm. The proposed protection scheme has been evaluated using simulated data obtained through EMTP/ATP package. The results amply demonstrate the capabilities of Fault Detector (FD) and Condition Monitor (CM) in terms of accuracy and speed with respect to detection of fault, classification, and pattern recognition of different events of power transformer.

Nonlinear channel equalization for QAM signal constellation using artificial neural networks

Application of artificial neural networks (ANN's) to adaptive channel equalization in a digital communication system with 4-QAM signal constellation is reported in this paper. A novel computationally efficient single layer functional link ANN (FLANN) is proposed for this purpose. This network has a simple structure in which the nonlinearity is introduced by functional expansion of the input pattern by trigonometric polynomials. Because of input pattern enhancement, the FLANN is capable of forming arbitrarily nonlinear decision boundaries and can perform complex pattern classification tasks. Considering channel equalization as a nonlinear classification problem, the FLANN has been utilized for nonlinear channel equalization. The performance of the FLANN is compared with two other ANN structures [a multilayer perceptron (MLP) and a polynomial perceptron network (PPN)] along with a conventional linear LMS-based equalizer for different linear and nonlinear channel models. The effect of eigenvalue ratio (EVR) of input correlation matrix on the equalizer performance has been studied. The comparison of computational complexity involved for the three ANN structures is also provided.

Editorial Artificial Neural Networks To Systems, Man, And Cybernetics: Characteristics, Structures, And Applications

The goal of this special issue is to present recent high- quality papers that deal with the applications of artificial neural networks (ANN's) to systems, man, and cybernetics (SMC). This special issue explores the state-of-the art in the applications of ANN to the SMC community. ANN's technology has reached a degree of maturity as evidenced by the increasing number of applications including the ones that have been reduced to practice. In this editorial, we present background and theoretical information related to ANN's, general characteristics, models, applications, and structures of ANN's. RTIFICIAL neural networks (ANN's) and neural en- gineering/computing in the wide sense are among to- day's most rapidly developing scientific disciplines. ANN's are parallel computational models that consist mainly of in- terconnected adaptive processing units. These networks are considered fine-grained parallel implementation of nonlinear dynamic and static systems. An ANN is an abstract simulation of real nervous system that contains a collection of processing units or processing elements (PE's) communicating with each other via axon connections. Such a model resembles the axons and dendrites of the nervous system. Because of its self- organizing and adaptive nature, the model provides a new parallel and distributed paradigm that has the potential to be more robust and user-friendly than traditional schemes (1)-(14). The study of artificial neural networks is an attempt to simulate and understand biological processes in an intrigu- ing manner. Today, we are witnessing the dawn of a new revolution in technology that will revamp the infrastructure of many approaches to solve cybernetics, information, and system engineering problems, among others. It is of interest to define alternative computational paradigms that attempt to mimic the brain's operation in several ways. Neural networks are an alternative approach to the traditional von Neumann programming schemes.